Multi-focal non-parallel collimator-based imaging

Abstract

A system and method include training of an artificial neural network to generate a simulated attenuation-corrected reconstructed volume from an input non-attenuation-corrected reconstructed volume, the training based on a plurality of non-attenuation-corrected volumes generated from respective ones of a plurality of sets of two-dimensional emission data and on a plurality of attenuation-corrected reconstructed volumes generated from respective ones of the plurality of sets of two-dimensional emission data.

What is claimed is: 
     
       1. A system comprising:
 a storage device; 
 a processor to execute processor-executable process steps stored on the storage device to cause the system to:
 generate a plurality of non-attenuation-corrected reconstructed volumes, each of the non-attenuation-corrected reconstructed volumes generated based on a respective one of a plurality of sets of two-dimensional emission data; 
 generate a plurality of attenuation-corrected reconstructed volumes, each of the attenuation-corrected reconstructed volumes generated based on a respective one of the plurality of sets of two-dimensional emission data; and 
 train an artificial neural network to generate a simulated attenuation-corrected reconstructed volume from an input non-attenuation-corrected reconstructed volume, the training based on the plurality of non-attenuation-corrected volumes and respective ones of the attenuation-corrected reconstructed volumes. 
 
 
     
     
       2. A system according to  claim 1 , wherein the artificial neural network is a convolutional network, and wherein the processor is to execute processor-executable process steps to cause the system to:
 output trained kernels of the trained network to an emission imaging system. 
 
     
     
       3. A system according to  claim 2 , further comprising the emission imaging system, the emission imaging system to:
 acquire a set of two-dimensional emission data; 
 reconstruct a non-attenuation-corrected volume based on the set of two-dimensional emission data; 
 input the non-attenuation-corrected reconstructed volume to a second convolutional network comprising the trained kernels; and 
 receive a simulated attenuation-corrected reconstructed volume generated by the second convolutional network based on the input non-attenuation-corrected reconstructed volume. 
 
     
     
       4. A system according to  claim 1 , further comprising an emission imaging system, the emission imaging system to:
 acquire a set of two-dimensional emission data using a multi-focal non-parallel collimator; 
 reconstruct a non-attenuation-corrected volume based on the set of two-dimensional emission data; 
 input the non-attenuation-corrected reconstructed volume to the trained network; and 
 receive a simulated attenuation-corrected reconstructed volume generated by the trained network based on the input non-attenuation-corrected reconstructed volume. 
 
     
     
       5. A system according to  claim 1 , wherein the processor is to execute processor-executable process steps to cause the system to:
 acquire a polar map associated with each of the non-attenuation-corrected reconstructed volumes, and wherein training the artificial neural network comprises: 
 training the artificial neural network to generate a simulated attenuation-corrected reconstructed volume from an input non-attenuation-corrected reconstructed volume and a polar map, the training based on the plurality of non-attenuation-corrected volumes and respective ones of the attenuation-corrected reconstructed volumes and the polar maps. 
 
     
     
       6. A system according to  claim 5 , wherein the processor is to execute processor-executable process steps to cause the system to:
 acquire an orbit length associated with each of the non-attenuation-corrected reconstructed volumes, and wherein training the artificial neural network comprises: 
 training the artificial neural network to generate a simulated attenuation-corrected reconstructed volume from an input non-attenuation-corrected reconstructed volume, a polar map and an orbit length, the training based on the plurality of non-attenuation-corrected volumes and respective ones of the attenuation-corrected reconstructed volumes, the polar maps and the orbit lengths. 
 
     
     
       7. A system according to  claim 1 , wherein the plurality of sets of two-dimensional emission data comprise SPECT data acquired using a multi-focal non-parallel collimator. 
     
     
       8. A method comprising:
 generating a plurality of non-attenuation-corrected reconstructed volumes, each of the non-attenuation-corrected reconstructed volumes generated based on a respective one of a plurality of sets of two-dimensional emission data; 
 generating a plurality of attenuation-corrected reconstructed volumes, each of the attenuation-corrected reconstructed volumes generated based on a respective one of the plurality of non-attenuation-corrected reconstructed volumes; and 
 training an artificial neural network to generate a simulated attenuation-corrected reconstructed volume from an input non-attenuation-corrected reconstructed volume, 
 the training based on the plurality of non-attenuation-corrected volumes and respective ones of the attenuation-corrected reconstructed volumes. 
 
     
     
       9. A method according to  claim 8 , wherein the artificial neural network is a convolutional network, and the method further comprising:
 outputting trained kernels of the trained network to an emission imaging system. 
 
     
     
       10. A method according to  claim 9 , further comprising;
 acquiring a set of two-dimensional emission data; 
 reconstructing a non-attenuation-corrected volume based on the set of two-dimensional emission data; 
 inputting the non-attenuation-corrected reconstructed volume to a second convolutional network comprising the trained kernels; and 
 receiving a simulated attenuation-corrected reconstructed volume generated by the second convolutional network based on the input non-attenuation-corrected reconstructed volume. 
 
     
     
       11. A method according to  claim 8 , further comprising:
 acquiring a set of two-dimensional emission data using a multi-focal non-parallel collimator; 
 reconstructing a non-attenuation-corrected volume based on the set of two-dimensional emission data; 
 inputting the non-attenuation-corrected reconstructed volume to the trained network; and 
 receiving a simulated attenuation-corrected reconstructed volume generated by the trained network based on the input non-attenuation-corrected reconstructed volume. 
 
     
     
       12. A method according to  claim 8 , further comprising:
 acquiring a polar map associated with each of the non-attenuation-corrected reconstructed volumes, and wherein training the artificial neural network comprises: 
 training the artificial neural network to generate a simulated attenuation-corrected reconstructed volume from an input non-attenuation-corrected reconstructed volume and a polar map, the training based on the plurality of non-attenuation-corrected volumes and respective ones of the attenuation-corrected reconstructed volumes and the polar maps. 
 
     
     
       13. A method according to  claim 12 , further comprising:
 acquiring an orbit length associated with each of the non-attenuation-corrected reconstructed volumes, and wherein training the artificial neural network comprises: 
 training the artificial neural network to generate a simulated attenuation-corrected reconstructed volume from an input non-attenuation-corrected reconstructed volume, a polar map and an orbit length, the training based on the plurality of non-attenuation-corrected volumes and respective ones of the attenuation-corrected reconstructed volumes, the polar maps and the orbit lengths. 
 
     
     
       14. A method according to  claim 8 , wherein the plurality of sets of two-dimensional emission data comprise SPECT data acquired using a multi-focal non-parallel collimator. 
     
     
       15. A system comprising:
 a storage device storing:
 a plurality of non-attenuation-corrected reconstructed volumes, each of the plurality of non-attenuation-corrected reconstructed volumes generated based on a respective one of a plurality of sets of two-dimensional emission data; 
 a plurality of attenuation-corrected reconstructed volumes, each of the attenuation-corrected reconstructed volumes generated based on a respective one of the plurality of sets of two-dimensional emission data; and 
 nodes of an artificial neural network; and 
 
 a processor to execute processor-executable process steps stored on the storage device to cause the system to:
 train the nodes of the artificial neural network to generate a simulated attenuation-corrected reconstructed volume from an input non-attenuation-corrected reconstructed volume, the training based on the plurality of non-attenuation-corrected volumes and respective ones of the plurality of attenuation-corrected reconstructed volumes. 
 
 
     
     
       16. A system according to  claim 15 , wherein the artificial neural network is a convolutional network, and wherein the processor is to execute processor-executable process steps to cause the system to:
 output trained kernels of the trained network nodes to an emission imaging system. 
 
     
     
       17. A system according to  claim 16 , further comprising the emission imaging system, the emission imaging system to:
 acquire a set of two-dimensional emission data; 
 reconstruct a non-attenuation-corrected volume based on the set of two-dimensional emission data; 
 input the non-attenuation-corrected reconstructed volume to a second convolutional network comprising the trained kernels; and 
 receive a simulated attenuation-corrected reconstructed volume generated by the second convolutional network based on the input non-attenuation-corrected reconstructed volume. 
 
     
     
       18. A system according to  claim 15 , further comprising an emission imaging system, the emission imaging system to:
 acquire a set of two-dimensional emission data using a multi-focal non-parallel collimator; 
 reconstruct a non-attenuation-corrected volume based on the set of two-dimensional emission data; 
 input the non-attenuation-corrected reconstructed volume to the trained network nodes; and 
 receive a simulated attenuation-corrected reconstructed volume generated by the trained network nodes based on the input non-attenuation-corrected reconstructed volume. 
 
     
     
       19. A system according to  claim 15 , the storage device to further store a polar map associated with each of the non-attenuation-corrected reconstructed volumes, and
 wherein training the nodes of the artificial neural network comprises training the nodes of the artificial neural network to generate a simulated attenuation-corrected reconstructed volume from an input non-attenuation-corrected reconstructed volume and a polar map, the training based on the plurality of non-attenuation-corrected volumes and respective ones of the attenuation-corrected reconstructed volumes and the polar maps. 
 
     
     
       20. A system according to  claim 19 , wherein the plurality of sets of two-dimensional emission data comprise SPECT data acquired using a multi-focal non-parallel collimator.